Unsupervised Domain Adaptation in LiDAR Semantic Segmentation with Self-Supervision and Gated Adapters
This addresses a practical problem for autonomous driving and robotics applications where LiDAR sensor variations cause model failures, though it appears incremental as it builds on existing domain adaptation techniques.
The paper tackles the problem of performance degradation in LiDAR semantic segmentation when models are tested on data from different sensors than they were trained on, proposing an unsupervised domain adaptation framework that achieves significant improvement over prior methods on real-to-real and synthetic-to-real benchmarks.
In this paper, we focus on a less explored, but more realistic and complex problem of domain adaptation in LiDAR semantic segmentation. There is a significant drop in performance of an existing segmentation model when training (source domain) and testing (target domain) data originate from different LiDAR sensors. To overcome this shortcoming, we propose an unsupervised domain adaptation framework that leverages unlabeled target domain data for self-supervision, coupled with an unpaired mask transfer strategy to mitigate the impact of domain shifts. Furthermore, we introduce the gated adapter module with a small number of parameters into the network to account for target domain-specific information. Experiments adapting from both real-to-real and synthetic-to-real LiDAR semantic segmentation benchmarks demonstrate the significant improvement over prior arts.